INDEX

Explanations

references to specific game numbers in sports

oai_token-act-pair · gpt-3.5-turbo

references to sporting events and game statistics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

anty

-0.81

company

-0.72

 skilled

-0.67

future

-0.67

hidden

-0.66

inet

-0.66

atan

-0.66

iw

-0.65

mor

-0.65

urat

-0.64

POSITIVE LOGITS

 festivities

0.74

 Arena

0.73

ħĭ

0.71

 onwards

0.71

 Tactics

0.71

 onward

0.68

ITIES

0.67

 Finals

0.66

 Countdown

0.66

 Premiere

0.65

Activations Density 0.056%

references to specific game numbers in sports

references to sporting events and game statistics

No Comments

No Known Activations