INDEX

Explanations

references to the physical sensation or state of having a hangover

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "over" in various contexts

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

ulet

-0.76

udes

-0.68

redits

-0.67

illary

-0.66

WI

-0.66

å¥³

-0.66

é¾

-0.65

ãĥ¼ãĥĨãĤ£

-0.64

orsi

-0.63

anyon

-0.63

POSITIVE LOGITS

lord

0.94

lords

0.88

leaf

0.86

ride

0.84

haul

0.83

loading

0.82

stable

0.80

flow

0.76

agus

0.76

due

0.76

Activations Density 0.012%

references to the physical sensation or state of having a hangover

occurrences of the word "over" in various contexts

No Comments

No Known Activations