INDEX

Explanations

terms related to reasons or purposes

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate reasons or justifications for actions or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

å§«

-0.72

ificant

-0.70

º

-0.66

ãģķ

-0.65

çĶ

-0.65

SN

-0.65

ãģª

-0.65

IME

-0.64

EStreamFrame

-0.63

IRC

-0.63

POSITIVE LOGITS

atics

0.70

hots

0.66

hops

0.66

 akin

0.65

 guiActiveUn

0.64

 reminiscent

0.61

 extraord

0.60

paces

0.59

vernight

0.59

uggest

0.59

Activations Density 0.433%

terms related to reasons or purposes

phrases that indicate reasons or justifications for actions or conditions

No Comments

No Known Activations