INDEX

Explanations

phrases describing qualities or characteristics of objects, products or events

oai_token-act-pair · gpt-3.5-turbo

repetitive references to the word "one."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

osponsors

-0.90

ories

-0.77

inders

-0.75

allas

-0.71

hips

-0.69

srf

-0.65

arlane

-0.65

untu

-0.64

ume

-0.64

ico

-0.63

POSITIVE LOGITS

 hundred

0.84

guy

0.82

 Hundred

0.81

 liner

0.80

 pesky

0.79

 thing

0.74

 bothered

0.73

 iteration

0.73

 sided

0.72

 dude

0.70

Activations Density 0.043%

phrases describing qualities or characteristics of objects, products or events

repetitive references to the word "one."

No Comments

No Known Activations