INDEX

Explanations

technical terms and specific document components like headings and captions

oai_token-act-pair · gpt-3.5-turbo

words or phrases related to specific actions or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

library

-0.69

avez

-0.68

laws

-0.67

nikov

-0.65

bury

-0.64

mere

-0.64

 pedigree

-0.63

tc

-0.62

lag

-0.61

 unaccount

-0.60

POSITIVE LOGITS

TION

1.00

 WATCHED

0.82

rontal

0.78

]}

0.74

luster

0.72

CHAT

0.71

 Fres

0.71

pmwiki

0.70

acular

0.69

 Frieza

0.68

Activations Density 0.007%

technical terms and specific document components like headings and captions

words or phrases related to specific actions or processes

No Comments

No Known Activations