INDEX

Explanations

phrases related to statements or comments made by individuals

oai_token-act-pair · gpt-3.5-turbo

references to comments or statements made by individuals

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

hess

-0.79

Eag

-0.67

 Shant

-0.67

SHA

-0.65

lance

-0.63

Pwr

-0.62

iage

-0.62

 Eucl

-0.60

 Tycoon

-0.60

 Everest

-0.60

POSITIVE LOGITS

 aloud

0.88

 uttered

0.79

 phrases

0.77

rompt

0.77

 speeches

0.74

plain

0.72

 poems

0.69

ettings

0.69

 louder

0.68

 slogans

0.68

Activations Density 0.629%

phrases related to statements or comments made by individuals

references to comments or statements made by individuals

No Comments

No Known Activations