INDEX

Explanations

names of popular news show programs

oai_token-act-pair · gpt-3.5-turbo

television network names and related broadcasting references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

alties

-0.73

orsche

-0.71

 robbers

-0.69

irez

-0.68

pur

-0.66

 starved

-0.66

raints

-0.65

adobe

-0.64

ËĪ

-0.63

sts

-0.63

POSITIVE LOGITS

 Tonight

1.11

 Colbert

0.96

CNN

0.95

 Hannity

0.92

Morning

0.91

Radio

0.91

Reilly

0.90

 MSNBC

0.90

Politics

0.87

CBS

0.86

Activations Density 0.224%

names of popular news show programs

television network names and related broadcasting references

No Comments

No Known Activations