INDEX

Explanations

words related to communication, specifically talking and discussions

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "talk."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

uilt

-0.81

necess

-0.79

 eleph

-0.78

ortunately

-0.78

emale

-0.77

 insign

-0.72

served

-0.72

boa

-0.70

ilitary

-0.69

aples

-0.68

POSITIVE LOGITS

 Talk

1.08

 talk

1.08

 aloud

0.85

talk

0.84

cloth

0.82

Talk

0.81

idian

0.80

ALK

0.77

tion

0.77

 talked

0.77

Activations Density 0.018%

words related to communication, specifically talking and discussions

occurrences of the word "talk."

No Comments

No Known Activations