INDEX

Explanations

keywords related to regulation, proposals, reports, accents, features, stimulus, policies, failures, mic, leftist frameworks, high-profile events, fights, crowds, competing studios, historical figures, protesters, gatherings, challenging species, moving individuals

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

catentry

-0.69

?".

-0.69

 Âł Âł

-0.67

ttp

-0.63

)</

-0.60

pc

-0.59

 ÂłÂł

-0.59

 english

-0.58

/?

-0.58

)\

-0.57

POSITIVE LOGITS

 chiefly

0.98

 insofar

0.96

 principally

0.90

 because

0.88

 precisely

0.84

 acutely

0.84

—

0.78

 primarily

0.77

—

0.76

 partly

0.76

Activations Density 1.033%

keywords related to regulation, proposals, reports, accents, features, stimulus, policies, failures, mic, leftist frameworks, high-profile events, fights, crowds, competing studios, historical figures, protesters, gatherings, challenging species, moving individuals

No Comments

No Known Activations