INDEX

Explanations

dates in the format "Month Day" (e.g., Sept 9)

oai_token-act-pair · gpt-3.5-turbo

dates, specifically those in September

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Reviewer

-0.75

jriwal

-0.71

 stricken

-0.69

culosis

-0.67

EStreamFrame

-0.66

Interstitial

-0.66

holder

-0.63

 reckless

-0.62

 Hots

-0.61

hyde

-0.59

POSITIVE LOGITS

Aug

0.86

Aug

0.86

0.84

alez

0.84

ober

0.82

Sept

0.82

 Sept

0.81

Oct

0.80

ibel

0.77

icka

0.76

Activations Density 0.004%

dates in the format "Month Day" (e.g., Sept 9)

dates, specifically those in September

No Comments

No Known Activations