INDEX

Explanations

expressions related to being in the process of doing something

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "about."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

gallery

-0.66

ATH

-0.65

rang

-0.65

 Peaks

-0.64

©¶æ¥µ

-0.64

ias

-0.63

rift

-0.62

NetMessage

-0.62

OGR

-0.62

Lens

-0.62

POSITIVE LOGITS

 halfway

0.88

 midway

0.67

 Ambro

0.62

 reforming

0.62

pload

0.61

 Ceres

0.60

 protecting

0.60

Activations Density 0.136%

expressions related to being in the process of doing something

occurrences of the word "about."

No Comments

No Known Activations