INDEX

Explanations

verbs related to actions, transformations, and modifications

oai_token-act-pair · gpt-3.5-turbo

actions related to change or modification

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 trailing

-0.66

 Suffolk

-0.64

 clipboard

-0.62

 gotta

-0.60

 bush

-0.59

 Hague

-0.58

Fuj

-0.57

 Horton

-0.56

 holders

-0.55

 Keller

-0.54

POSITIVE LOGITS

 safely

0.91

 anytime

0.90

 easily

0.84

 cheaply

0.83

 offline

0.79

 anywhere

0.76

 manually

0.76

 indefinitely

0.75

 securely

0.74

 separately

0.74

Activations Density 0.136%

verbs related to actions, transformations, and modifications

actions related to change or modification

No Comments

No Known Activations