INDEX

Explanations

instances of the word "supported"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "supported."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

odor

-0.73

iter

-0.72

azard

-0.72

oufl

-0.69

Rin

-0.69

blem

-0.69

hunt

-0.69

vent

-0.67

awk

-0.67

kay

-0.67

POSITIVE LOGITS

 supported

1.13

 Supported

0.99

 supports

0.94

 Supports

0.93

 support

0.87

 supporting

0.87

supported

0.85

Supported

0.81

 corrobor

0.80

Support

0.80

Activations Density 0.008%

instances of the word "supported"

instances of the word "supported."

No Comments

No Known Activations