INDEX

Explanations

information or answers by actively seeking them out through investigation or inquiry

oai_token-act-pair · gpt-3.5-turbo

verbs related to inquiry or discovery

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

isSpecialOrderable

-0.69

Reviewer

-0.66

 ISBN

-0.64

 depends

-0.63

 stumble

-0.62

 Subtle

-0.61

wikipedia

-0.61

 Thumbnails

-0.61

urses

-0.60

iasco

-0.60

POSITIVE LOGITS

 feasibility

0.75

 authenticity

0.75

uate

0.74

 compliance

0.68

 constituents

0.67

ether

0.67

 inspir

0.65

RM

0.63

 conformity

0.63

 possible

0.62

Activations Density 0.265%

information or answers by actively seeking them out through investigation or inquiry

verbs related to inquiry or discovery

No Comments

No Known Activations