INDEX

Explanations

specific mentions of news channels like ABC, CBS, Fox, CNN, and NBC in interviews and news segments

oai_token-act-pair · gpt-3.5-turbo

references to television networks and shows

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

gra

-0.80

abiding

-0.77

wikipedia

-0.74

origin

-0.73

vironment

-0.71

artist

-0.70

ancial

-0.69

pattern

-0.67

uscript

-0.66

ims

-0.66

POSITIVE LOGITS

 interviewer

1.12

 CNBC

0.90

 Hannity

0.87

 aired

0.84

CNN

0.83

 interview

0.83

 microphones

0.80

 interviews

0.79

 MSNBC

0.79

 Kimmel

0.76

Activations Density 0.152%

specific mentions of news channels like ABC, CBS, Fox, CNN, and NBC in interviews and news segments

references to television networks and shows

No Comments

No Known Activations