INDEX

Explanations

words related to unauthorized departures or changes in allegiance

oai_token-act-pair · gpt-3.5-turbo

terms related to detection and sectors in various contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

pled

-0.89

ples

-0.86

ãĥ£

-0.79

hens

-0.74

fulness

-0.74

mob

-0.73

thia

-0.69

wal

-0.68

trans

-0.68

pler

-0.66

POSITIVE LOGITS

naires

0.87

ary

0.85

naire

0.83

alid

0.76

aries

0.74

ection

0.71

nel

0.70

arians

0.70

eper

0.68

yne

0.67

Activations Density 0.029%

words related to unauthorized departures or changes in allegiance

terms related to detection and sectors in various contexts

No Comments

No Known Activations