INDEX

Explanations

terms related to improvement or enhancement

oai_token-act-pair · gpt-3.5-turbo

terms related to improvement or enhancement

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

PsyNetMessage

-0.74

 SHARES

-0.63

 Provided

-0.62

lication

-0.62

 answers

-0.60

 ACTIONS

-0.59

rities

-0.59

!/

-0.59

 Niet

-0.59

 resolutions

-0.58

POSITIVE LOGITS

cone

0.83

 richer

0.77

 worthwhile

0.74

itably

0.67

rongh

0.67

habi

0.64

readable

0.64

etitive

0.64

 tame

0.64

vind

0.64

Activations Density 0.202%

terms related to improvement or enhancement

terms related to improvement or enhancement

No Comments

No Known Activations