INDEX

Explanations

names of individuals or groups, typically emphasizing their roles or actions

oai_token-act-pair · gpt-3.5-turbo

specific names and references related to individuals or groups involved in negotiations or conflicts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

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

multiple

-0.72

sample

-0.70

dimension

-0.67

izons

-0.66

¥µ

-0.66

ħĭ

-0.66

"],"

-0.65

_>

-0.64

etheless

-0.63

ļéĨĴ

-0.63

POSITIVE LOGITS

 fault

1.19

 alone

0.82

 versus

0.76

 Fault

0.75

vs

0.72

who

0.67

 coincidence

0.66

 responsibility

0.66

 raining

0.66

hest

0.66

Activations Density 0.278%

names of individuals or groups, typically emphasizing their roles or actions

specific names and references related to individuals or groups involved in negotiations or conflicts

No Comments

No Known Activations