INDEX

Explanations

words related to legal matters and law enforcement

oai_token-act-pair · gpt-3.5-turbo

punctuated lists or series of items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ocl

-0.75

²¾

-0.75

estones

-0.72

rush

-0.71

lihood

-0.71

enth

-0.70

uten

-0.67

cloth

-0.65

sburg

-0.64

è¦ļéĨĴ

-0.64

POSITIVE LOGITS

 according

0.90

 prompting

0.86

 CNBC

0.81

 POLITICO

0.81

 sparking

0.80

AFP

0.73

CBC

0.72

 despite

0.70

 claiming

0.70

 confirming

0.70

Activations Density 0.283%

words related to legal matters and law enforcement

punctuated lists or series of items

No Comments

No Known Activations