INDEX

Explanations

The neuron is looking for mentions of legal terms and institutions

oai_token-act-pair · gpt-3.5-turbo

the letter 'a' in various contexts throughout the document

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

lasses

-0.74

ARGET

-0.70

ymm

-0.68

ONSORED

-0.66

<[

-0.65

inances

-0.65

pter

-0.64

 WATCHED

-0.64

rored

-0.64

onel

-0.62

POSITIVE LOGITS

Ã±

1.15

qua

0.97

Ã°

0.96

ption

0.95

Ã±a

0.92

BILITY

0.91

vel

0.90

ichi

0.89

emia

0.89

ishi

0.87

Activations Density 0.071%

The neuron is looking for mentions of legal terms and institutions

the letter 'a' in various contexts throughout the document

No Comments

No Known Activations