INDEX

Explanations

legal and criminal terms related to obstruction

oai_token-act-pair · gpt-3.5-turbo

terms related to legal obstruction and interference

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

ndra

-0.88

ann

-0.78

mont

-0.76

morph

-0.71

 Stras

-0.71

Fly

-0.68

variable

-0.68

 Grind

-0.68

mut

-0.67

MON

-0.67

POSITIVE LOGITS

 obstruction

1.25

 obstruct

1.10

structed

0.90

acles

0.83

 imped

0.81

struction

0.81

 impede

0.78

uments

0.77

ions

0.76

 overshadow

0.75

Activations Density 0.023%

legal and criminal terms related to obstruction

terms related to legal obstruction and interference

No Comments

No Known Activations