INDEX

Explanations

statements indicating who said a particular phrase

oai_token-act-pair · gpt-3.5-turbo

occurrences of the verb "says" or synonymous terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 categor

-0.78

 grav

-0.73

 underpin

-0.66

 shut

-0.65

 hyper

-0.65

 climax

-0.64

 balance

-0.63

 remain

-0.63

 duel

-0.62

 fluct

-0.61

POSITIVE LOGITS

 Says

2.03

 Said

2.02

 Gets

2.00

 Makes

1.89

 Doesn

1.89

 Appears

1.86

 Turns

1.84

 Didn

1.83

 Called

1.83

 Seems

1.83

Activations Density 0.052%

statements indicating who said a particular phrase

occurrences of the verb "says" or synonymous terms

No Comments

No Known Activations