INDEX

Explanations

adjectives related to intensity or focus

oai_token-act-pair · gpt-3.5-turbo

terms related to politicized and emotionally charged topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

redits

-0.82

uther

-0.82

registered

-0.80

verified

-0.76

âĿ

-0.75

chwitz

-0.75

drug

-0.75

 besides

-0.74

Specific

-0.73

Film

-0.73

POSITIVE LOGITS

 nature

1.19

 confines

0.94

 portion

0.90

 assumption

0.87

 notion

0.87

 aspect

0.85

 process

0.83

 version

0.82

 phase

0.82

 tendency

0.81

Activations Density 0.313%

adjectives related to intensity or focus

terms related to politicized and emotionally charged topics

No Comments

No Known Activations