INDEX

Explanations

mentions of sports qualifiers or qualifying games

oai_token-act-pair · gpt-3.5-turbo

mentions of the term "qualifying" 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 @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

orah

-0.82

 Kinnikuman

-0.79

 Strauss

-0.72

ographed

-0.71

 Rider

-0.70

corn

-0.70

":"/

-0.69

RH

-0.67

relative

-0.65

utra

-0.64

POSITIVE LOGITS

 qualifying

0.90

 qualifiers

0.84

 qualifications

0.80

 qualify

0.80

ifying

0.71

 qualification

0.71

ifiers

0.70

idad

0.67

ifies

0.65

 qualified

0.65

Activations Density 0.006%

mentions of sports qualifiers or qualifying games

mentions of the term "qualifying" in various contexts

No Comments

No Known Activations