INDEX

Explanations

The neuron looks for phrases indicating a recurring action or event

oai_token-act-pair · gpt-3.5-turbo

the phrase "once again."

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

Runner

-0.65

osit

-0.63

 Treaty

-0.63

ĺħ

-0.62

otti

-0.61

OTS

-0.59

zie

-0.59

ohm

-0.58

rament

-0.58

adena

-0.58

POSITIVE LOGITS

forth

0.91

nir

0.81

nces

0.73

brates

0.72

RGB

0.71

 oblig

0.71

theless

0.70

brate

0.69

 conclud

0.68

atility

0.65

Activations Density 0.022%

The neuron looks for phrases indicating a recurring action or event

the phrase "once again."

No Comments

No Known Activations