INDEX

Explanations

phrases where someone is expressing an opinion or argument

oai_token-act-pair · gpt-3.5-turbo

statements or claims made by individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ptives

-0.72

adesh

-0.70

theless

-0.65

================================================================

-0.63

prus

-0.63

ãĥ¼ãĥĨ

-0.62

estern

-0.60

https

-0.60

uador

-0.60

FTWARE

-0.60

POSITIVE LOGITS

,,

0.76

*,

0.72

0.71

 convinc

0.70

 bluntly

0.66

 goodbye

0.61

!,

0.60

 omin

0.59

cyn

0.59

 emphatically

0.58

Activations Density 0.187%

phrases where someone is expressing an opinion or argument

statements or claims made by individuals

No Comments

No Known Activations