INDEX

Explanations

statements involving past actions or outcomes

oai_token-act-pair · gpt-3.5-turbo

phrases related to community involvement and shared experiences

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

 Nato

-0.70

 NATO

-0.65

 Hezbollah

-0.64

elfare

-0.62

nsic

-0.61

 Reagan

-0.60

 evacuated

-0.60

 Workers

-0.60

 withdrawn

-0.59

 Saddam

-0.59

POSITIVE LOGITS

 guiActiveUnfocused

0.76

 haha

0.75

 fanbase

0.74

 fruition

0.72

;)

0.69

XD

0.69

acly

0.67

 hopefully

0.67

ãĤ´ãĥ³

0.66

:)

0.65

Activations Density 2.816%

statements involving past actions or outcomes

phrases related to community involvement and shared experiences

No Comments

No Known Activations