INDEX

Explanations

references to the word ''Ice''

oai_token-act-pair · gpt-3.5-turbo

references to "Ice" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

verb

-0.83

________________

-0.79

=-=-=-=-=-=-=-=-

-0.79

edy

-0.78

rompt

-0.77

 candidates

-0.75

pled

-0.72

INGTON

-0.71

anooga

-0.70

atche

-0.69

POSITIVE LOGITS

Ice

1.12

breaker

1.02

 rink

0.98

 Cream

0.95

breakers

0.93

 cream

0.89

 Hockey

0.83

 skating

0.80

Age

0.80

Ice

0.78

Activations Density 0.006%

references to the word ''Ice''

references to "Ice" in various contexts

No Comments

No Known Activations