INDEX

Explanations

mentions of the word "telegram."

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "telegram" and its variations

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

 Augustine

-0.69

cca

-0.67

package

-0.65

sharing

-0.65

 PLAN

-0.63

 Angelo

-0.62

GD

-0.61

const

-0.60

loving

-0.60

 Ambrose

-0.60

POSITIVE LOGITS

legraph

1.38

orld

0.92

itudinal

0.85

kefeller

0.84

biz

0.83

ombat

0.81

emouth

0.78

ijk

0.77

htaking

0.76

hedon

0.76

Activations Density 0.006%

mentions of the word "telegram."

occurrences of the word "telegram" and its variations

No Comments

No Known Activations