INDEX

Explanations

verbs related to tasks or actions

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "doing" in various contexts

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

lights

-0.73

Tier

-0.71

mares

-0.70

Tai

-0.69

mare

-0.67

rum

-0.66

dayName

-0.66

tight

-0.65

 Handling

-0.65

pa

-0.63

POSITIVE LOGITS

omething

0.88

omsday

0.86

pez

0.82

 mosqu

0.81

oley

0.79

berman

0.76

lyak

0.76

 something

0.74

 nothing

0.74

ggy

0.73

Activations Density 0.035%

verbs related to tasks or actions

occurrences of the word "doing" in various contexts

No Comments

No Known Activations