INDEX

Explanations

inflammatory terms related to medical conditions or social situations

oai_token-act-pair · gpt-3.5-turbo

terms associated with inflammation and related medical conditions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ģĸ

-0.81

ibling

-0.77

hare

-0.77

fman

-0.74

etheus

-0.74

ILY

-0.73

ault

-0.71

fell

-0.71

ille

-0.71

awei

-0.69

POSITIVE LOGITS

 cytok

1.07

 bowel

0.92

 arthritis

0.90

 inflammatory

0.90

 infiltrate

0.88

lamm

0.87

 inflammation

0.84

inflammatory

0.84

 inflamm

0.79

 Tend

0.72

Activations Density 0.018%

inflammatory terms related to medical conditions or social situations

terms associated with inflammation and related medical conditions

No Comments

No Known Activations