INDEX

Explanations

words related to negative or critical commentary

oai_token-act-pair · gpt-3.5-turbo

terms associated with intense visual or metaphorical brightness and severity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

icles

-0.82

eva

-0.81

udo

-0.80

xy

-0.80

icals

-0.78

istry

-0.77

odox

-0.76

sung

-0.75

icle

-0.74

ascript

-0.73

POSITIVE LOGITS

tons

0.79

 Archdemon

0.75

HAM

0.75

redients

0.73

 tides

0.71

 alcoholic

0.71

 oceans

0.71

 cess

0.70

 noises

0.68

BuyableInstoreAndOnline

0.68

Activations Density 0.086%

words related to negative or critical commentary

terms associated with intense visual or metaphorical brightness and severity

No Comments

No Known Activations