INDEX

Explanations

locations or activities related to physical exercise

oai_token-act-pair · gpt-3.5-turbo

references to gyms and fitness-related activities

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

 Strait

-0.82

drawn

-0.68

theless

-0.66

 afar

-0.65

ICT

-0.64

better

-0.64

Boe

-0.63

IBLE

-0.62

 flawed

-0.62

angered

-0.62

POSITIVE LOGITS

nas

1.60

rats

0.89

rosse

0.87

bell

0.85

cers

0.84

rition

0.82

nos

0.82

mers

0.79

 lapt

0.78

gym

0.77

Activations Density 0.011%

locations or activities related to physical exercise

references to gyms and fitness-related activities

No Comments

No Known Activations