INDEX

Explanations

instructions or guides with numbered steps

oai_token-act-pair · gpt-3.5-turbo

expressions of intention or plans for future actions

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

 attest

-0.75

soever

-0.75

often

-0.68

 whatsoever

-0.65

habi

-0.64

hovah

-0.64

 testified

-0.62

lees

-0.61

 jails

-0.61

clips

-0.59

POSITIVE LOGITS

 Kickstarter

0.73

 postpone

0.71

 nominate

0.70

 misunder

0.67

escription

0.65

clusion

0.65

 giveaway

0.65

ossibility

0.64

 myself

0.64

SELECT

0.64

Activations Density 0.562%

instructions or guides with numbered steps

expressions of intention or plans for future actions

No Comments

No Known Activations