INDEX

Explanations

words related to specifics or specific details

oai_token-act-pair · gpt-3.5-turbo

phrases that introduce detailed explanations or clarifications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 farewell

-0.66

 waiting

-0.65

 behind

-0.63

 goodbye

-0.63

 instinct

-0.62

urus

-0.62

 never

-0.62

 rushing

-0.62

no

-0.61

 twilight

-0.60

POSITIVE LOGITS

 Specifically

3.37

Specifically

3.07

 Notably

1.64

 Particularly

1.63

 Accordingly

1.49

 Essentially

1.42

 specifically

1.38

 Basically

1.35

Specific

1.31

 Included

1.30

Activations Density 0.014%

words related to specifics or specific details

phrases that introduce detailed explanations or clarifications

No Comments

No Known Activations