INDEX

Explanations

action words related to different tasks or actions

oai_token-act-pair · gpt-3.5-turbo

verbs related to actions that improve understanding or emphasize clarity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

+.

-0.59

inals

-0.59

usercontent

-0.56

 Travels

-0.53

"></

-0.52

WHERE

-0.52

 UNCLASSIFIED

-0.51

ãĥĲ

-0.51

sed

-0.51

 slashed

-0.51

POSITIVE LOGITS

 this

0.72

 analogy

0.71

 yourself

0.70

 myself

0.70

 matters

0.69

 properly

0.67

 nutshell

0.66

 your

0.65

uate

0.65

 ourselves

0.65

Activations Density 0.149%

action words related to different tasks or actions

verbs related to actions that improve understanding or emphasize clarity

No Comments

No Known Activations