INDEX

Explanations

instances where 'NC' followed by a number is mentioned

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of "NC," likely referring to a specific organization or entity

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

thing

-0.89

chest

-0.88

loo

-0.79

tons

-0.76

shake

-0.72

gerald

-0.71

nings

-0.71

pipe

-0.69

many

-0.67

downs

-0.65

POSITIVE LOGITS

urses

1.04

ODE

0.88

STAR

0.81

ouri

0.81

ossus

0.79

ategor

0.78

ADE

0.78

NC

0.77

urrent

0.77

ertain

0.75

Activations Density 0.009%

instances where 'NC' followed by a number is mentioned

repeated mentions of "NC," likely referring to a specific organization or entity

No Comments

No Known Activations