INDEX

Explanations

phrases related to self-identity and self-description

oai_token-act-pair · gpt-3.5-turbo

phrases associated with self-identification or self-description

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ulhu

-1.15

 "$:/

-0.79

 Chains

-0.71

 Hutch

-0.70

 Boone

-0.69

 Twain

-0.68

OUGH

-0.68

Cah

-0.68

 Springs

-0.68

 EntityItem

-0.67

POSITIVE LOGITS

talk

1.12

imposed

1.09

conscious

1.09

esteem

1.04

destruct

1.01

generation

1.00

eating

1.00

proclaimed

0.98

assert

0.96

expression

0.96

Activations Density 0.043%

phrases related to self-identity and self-description

phrases associated with self-identification or self-description

No Comments

No Known Activations