INDEX

Explanations

references to specific rankings or numbers

oai_token-act-pair · gpt-3.5-turbo

references to rankings and numbered designations in sports

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

ventions

-0.74

ework

-0.73

iths

-0.70

sym

-0.69

acci

-0.68

acc

-0.67

Ã©e

-0.65

rawdownloadcloneembedreportprint

-0.63

apters

-0.63

 recount

-0.63

POSITIVE LOGITS

 seed

0.83

ashington

0.81

 overall

0.75

 ranked

0.75

 seeds

0.74

 spot

0.74

 jersey

0.73

CB

0.70

 contender

0.69

CLASSIFIED

0.68

Activations Density 0.042%

references to specific rankings or numbers

references to rankings and numbered designations in sports

No Comments

No Known Activations