INDEX

Explanations

mentions or references to specific terms or topics within a longer text

oai_token-act-pair · gpt-3.5-turbo

references to previous statements or mentions in a discussion

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯

-0.93

Â¯Â¯

-0.76

Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯

-0.73

sis

-0.72

heric

-0.72

sett

-0.71

ccording

-0.70

Â¯Â¯Â¯Â¯

-0.70

zers

-0.70

ï¸

-0.69

POSITIVE LOGITS

 mentioning

0.94

 mentions

0.91

lihood

0.82

 Kislyak

0.81

 mention

0.76

 aloud

0.74

enance

0.73

 prominently

0.66

cliffe

0.64

 Vaughn

0.63

Activations Density 0.049%

mentions or references to specific terms or topics within a longer text

references to previous statements or mentions in a discussion

No Comments

No Known Activations